Towards Holistic Evaluation of Education Systems

Abstract

Children internationally are entitled to quality education. International education initiatives monitor education system quality through complex evaluations, historically relying on academic benchmarks operationalized by robust comparative achievement data. However, quality in schooling is evolving to comprise development beyond academic abilities – it should support emotional, social, and psychological development. Valid systems-level evaluations of these features internationally require well-defined benchmarks for school conditions suitably supportive of this development. Emerging international initiatives, such as UNESCO’s Happy Schools Framework, define frameworks for non-academic facets but have not been empirically tested. This study defines school climate health as the intersection of the Happy Schools Framework, existing literature on school climate and wellbeing, and the Trends in International Mathematics and Science (TIMSS) Context Questionnaire. This research aims to provide a first step towards defining benchmarks by exploring an international dataset to define existing patterns of interrelated school context variables. This study is responsive to empirical literature and relevant theoretical frameworks for evaluating social systems (systems evaluation, ecological systems theories). An exploratory multilevel latent class analysis (MLCA) of 22 variables is conducted for the 58 participating countries to define four school clusters and three country classes defining the composition and distribution of school climate health internationally. Combining response variables from students, teachers, principals, and parents is a novel application. Characteristics of each school cluster and country class are described. Secondary analyses investigate possible confoundedness of school demographics and possible relationships between school-level average achievement.